Job mismatches in Pakistan: is there some wage penalty to graduates?
Farooq, Shujaat
In this study, an attempt has been made to estimate the incidence
of job mismatch and its impacts on graduate's earnings in Pakistan.
The study has divided the job mismatch into three categories;
qualification-job mismatch, skill mismatch and field of study and job
mismatch. The primary dataset has been used in which the formal sector
employed graduates have been studied. This study has measured the
qualification-job mismatch by three approaches and found that about
one-third of the graduates are facing qualification-job mismatch.
Similarly, more than one-fourth of the graduates are mismatched in
skill, about half of them are over-skilled and the half are
under-skilled. The analysis also shows that 11.3 percent of the
graduates have irrelevant and 13.8 percent have slightly relevant jobs
to their studied field of disciplines. Our analysis shows that
over-qualified graduates face wage penalty under different approaches.
After controlling skill heterogeneity, there is less penalty to
apparently overqualified and more penalty to genuinely over-qualified.
The over-skilled graduates face wage penalties and the under-skilled get
wage premiums as compared to the matched workers. A good field of study
and job matches also improve the wages of graduates.
JEL Classification: 123, 124, J21, J24, J31
Keywords: Education and Inequality, Higher Education, Human
Capital, Labour Market, Wages
1. INTRODUCTION
The role of human capital has long been acknowledged by researchers
and policy makers not only for sustained economic growth but also for
social cohesion. Being so important, the policy-makers all around the
globe have stressed allocating more resources to raise education level,
which in turn, affects worker's earning and national productivity.
In 1960s and 70s, many developed countries including U.S and U.K started
to invest heavily in higher education, and Freeman (1976) was the first
who raised his concern while analysing the accuracy of the match between
graduates' attained education and education demanded by the labour
market. The initial studies perceived it as a temporary phenomenon
[Freeman (1976)]; however, it was not empirically supported as the
incidence of 'over-education', mainly focused on literature,
ranges from 10 percent to 40 percent, an average of 25 percent in
developed countries [Groot and Maassen (2000); Leuven and Oosterbeek
(2011)]. These estimates raised serious questions over the validity of
conventional views of the labour market; consequently a good debate has
started with the emergence of some new theories i.e. the job competition
theory and the job assignment theory in which the institutional
rigidities, allocation problems and skill heterogeneities were dealt.
Both the economists and sociologists have consigned the job
mismatch phenomenon as a serious efficiency concern with its pertinent
socio-economic costs at individual, firm and national level. At
individual level, it would decrease the individual's marginal
product as the existing studies show that over-qualified workers earn
less than the matched workers, though the estimated wage differentials
differ across the countries. (1) The lower returns to education may also
incur some non-transitory costs i.e. lower level of job satisfaction,
frustration and higher turnover rate. At the firm level, job mismatch is
associated with lower productivity and lower level of job involvement;
and in case of high turnover rates, firms may have to incur extra costs
on screening, recruiting and training [Tsang (1987); Sloane, et al.
(1999)]. At the macro level, the national welfare would be lowered by
under-utilisation of skills [McGuinnes (2006)]. It is also possible that
previously well-matched graduates in the economy will be 'bumped
down' in the labour market as over-qualified workers move into
lower occupations thus raising the educational requirements within these
occupations [Battu, et al. (2000)].
The phenomenon can be perceived from some studies, which have
highlighted educated unemployment and under-employment [Ghayur (1989);
Pakistan (2013)], skill heterogeneity due to educational expansion
[Haque, et al. (2007)] and decline in rate of return to education
[Hausman, et al. (2005); Qayyum, et al. (2007)]. Recently some studies
have emphasised this phenomenon in the context of role of education in
career development [Zahid (2014)]. The ongoing demographic transition in
Pakistan may also cause the job mismatch phenomenon as the labour force
grows faster than the employment rate. As a result, the quality of jobs
and access to modest earning opportunities has been emerging as a key
issue as reflected by the various labour indicators e.g. educated
unemployment, decline in worker's productivity, rising share of
informal labour, rising job search periods and high risk of
vulnerability especially for youths and females [Pakistan (2008,
2011,2013)]. (2)
Becker's (1964) monogram 'Human Capital' provides
the basic foundations to explain earning distribution in developed
countries and Mincer's model (1974) on earning provides a
cornerstone empirical framework to predict the human capital theory.
Both Becker (1964) and Mincer (1974) asserted that education and
training are the most important components of human capital
accumulation, which in turn, directly and indirectly affect the
individuals' life time earnings. Following Becker's Human
Capital Theory (1964), a number of studies in Pakistan have measured the
return to education by assuming that labour market is competitive and
workers are paid according to their marginal product. (3) But no study
has anticipated the impact of job mismatch on earnings. In view of the
importance of job mismatch and existing literature gap in Pakistan, the
study aims to measure the potential impact of various types of job
mismatchs on graduates' earning in Pakistan. Since terms
'education and job mismatch' are linked with educated workers,
therefore the analysis in this study is carried out on employed
graduates working in the formal sector who hold at least fourteen years
formal education, named as the 'graduate workers'.
The rest of the study is organised as follows. Section 2 presents
the theoretical framework of job mismatch discussing both: the types of
job mismatch and theoretic aspects of job mismatch. Discussion on data
sources and methodology is given in Section 3. The penultimate section
has discussed the results over the incidence of job mismatch and its
impact on graduate's earning. Conclusions and policy considerations
are given in the final section.
2. JOB MISMATCH AND WORKER'S EARNING: A THEORETICAL FRAMEWORK
Job mismatch has three dimensions; qualification-job mismatch,
skill mismatch and field of study and job mismatch [Farooq (2011)].
qualification-job mismatch compares the acquired qualification (in
years) with the required qualification (in years) of a worker in his/her
current job, while the skill mismatch compares overall acquired
competences with the required competences. The field of study and job
mismatch evaluates that how much studied field of discipline is relevant
to the nature of job. An extensive literature exists on the first type
of job mismatch; whereas, only few subjective studies recently have been
made on skill mismatch and field of study and job mismatch. All these
studies have been carried out primarily in the developed economies. The
existing studies are mixed over the use of titles for three types of job
mismatches as some studies have used the term 'qualification
mismatch' by Green and McIntosh (2002), and 'education
mismatch' by Verdugo and Verdugo, (1989), Battu, et al. (2000),
Lourdes, et al. (2005) etc. for the first type of job mismatch
(qualification mismatch). Similarly, different titles have been used for
the second type of job mismatch (skill mismatch) i.e. competence
mismatch by Lourdes, et al. (2005) and skill mismatch by Green and
McIntosh (2002), Jim and Egbert (2005) and Lourdes and Luis (2013).The
rest of this study will follow the titles as given in Figure 1;
qualification-job mismatch, skill mismatch and field of study and job
mismatch. The sub-classification of graduates under each type of job
mismatch is also given in Figure 1.
[FIGURE 1 OMITTED]
Though there is no unified accepted theory on job mismatch and
earnings; however, the following three theories have explained the job
mismatch phenomenon with earnings. According to Human Capital Theory
(HCT), labour market is competitive where every worker is paid the value
of his/her marginal product [Schultz (1962); Becker (1964)]. Wages and
productivity are fixed in relation to prospective jobs; therefore,
overqualified workers have same productivity and thus receive the same
wages as compared to the matched workers. In a pure human capital
framework, the concept of job mismatch may be meaningless. The job
mismatch phenomenon may not necessarily reject the HCT in case of short
run existence; however, if it appears to be a long run phenomenon, then
no one can save the HCT [McGuiness (2006)]. The opponents of HCT argue
that it fails to explain the underutilisation of skills, institutional
rigidities and non-competitive labour market. Tsang (1987) suggested
that the relationship between education and productivity is more
multifaceted than the direct and positive relationship as suggested by
HCT. Some studies have pointed out that return to education may not
increase with the level of education [World Bank in "Knowledge for
Development" (1999); Psacharopoulos and Patrinos (2002)].
In contrast to HCT, the Job Competition Theory highlights the
institutional rigidities where earnings are associated with job
characteristics [Thurow (1975)]. The allocation on job is based on
available supplies of both workers and jobs, workers may possess more
education and skills than their jobs necessitate. If there is an
over-supply of educated job seekers, some educated workers will look for
jobs at lower level with wage penalties. In the extreme case, education
simply serves to obtain the job, and there is a zero return to human
capital beyond that required to do the job. Therefore, Mincer model
(1974) and the Thurow's model (1975) are two extreme cases, the
first being purely supply side driven and the second being purely demand
side driven.
A third strand between the former two extreme cases is found in the
Job Assignment Theory, which asserts that there is an allocation problem
in assigning the heterogeneous workers to jobs which differ in their
complexity [Sattinger (1993)]. Hartog (2000) viewed that the labour
market is consisting of a bundle of capabilities and suggested that up
to 40 percent of the income variance can be attributed to capability
variables. In practice, the frequency distributions are unlikely to
match and education mismatch may be a persistent problem if the job
structure is relatively unresponsive to changes in relative supplies of
educated labour. Earnings are then a function of both individual and job
characteristics where over-qualified workers earn some rate of return on
over-education but less than the return to required education.
Duncan and Hoffman (1981) found that over-qualified workers receive
a lower return on surplus schooling. In Europe, similar findings have
been reported by Dolton and Vignoles (2000), Groot and Maasen (2000),
Battu, et al. (1999) and many others. A dominant paradigm of literature
concludes that over-qualified workers face wage penalties, while
under-qualified workers enjoy wage premiums while comparing them with
the matched workers with the same level of formal education. Initially,
these finding were reported by Verdugo and Verdugo (1989), Gill and
Solberg (1992). Later these results were endorsed by Cohn and Khan
(1995), Dolton and Vignoles (2000), Bauer (2002) and Frenette (2004).
The second finding is that the job mismatch explains the wage
differentials among workers who hold the same type of jobs. Thus, the
workers earn a positive rate of return on years of over-education, which
is lower than the required education (in years). Similarly,
under-qualified workers have a negative rate of return. These results
were initially estimated by Duncan and Hoffman (1981) and later
confirmed by Alba (1993), Sloane, et al. (1999), Groot and Maasen
(2000), Ng (2001), Groeneveld and Hartog (2004). Overall, the literature
supports the assignment theory that the over-qualified workers are
working below their potential but gaining some benefit from surplus
schooling [Alba (1993); Groot (1996); Sloane, et al. (1999); Hartog
(2000); Dolton and Silles (2003); Lourdes, et al. (2005); Chevalier and
Lindley (2006); Martin, et al. (2008)].
3. DATA SOURCE AND METHODOLOGY
3.1. Data Description
Due to non-availability of key information in national secondary
data sources including e.g. required education for a specific job,
attained and required level of skills, relevance of field of study to
current job and job satisfaction, the present study has used the primary
dataset by targeting the employed graduates working in the formal sector
who have fourteen and above years of education (Graduates, Master,
MS/MPhil, PhD), named as 'graduate workers'. A primary survey,
the Survey of Employed Graduates (SEG) has been conducted in 2010 in two
major cities of Pakistan, Islamabad and Rawalpindi to study the job
mismatch phenomenon in depth. At broad level, the targeted universe in
the SEG dataset has been divided into the three major groups; graduates
in federal government, graduates in autonomous/semi-autonomous bodies
under federal government and graduates in the private sector. The
Thirteenth Census Report of Federal Government Civil Servants (2003-04)
(4) and Annual Statistical Bulletin of Federal Government and
Semi-government (2007-08) (5) were used to estimate the graduate
employees in the federal government and semi-government. For private
sector, the relevant information was gathered from a few private
departments i.e. banks, hotels, telecom companies, international donor
offices, media (newspaper and broadcasting). For the remaining private
sector like hospitals, educational institutions, NGOs, manufacturing and
Industry etc., the internet and the other sources were used to get the
total numbers of units located in Islamabad/Rawalpindi and then through
rapid sample survey, the information was obtained to estimate the
employed graduates.
To avoid the sampling bias and errors, the proportional stratified
random sampling technique was adopted where the published BPS grades for
the government and semi-government sectors have been considered as
'strata' while the 3-digit occupational codes were used as
'strata' for the private sector. For further detail on
population universe and sampling, see Farooq (2011). A sample of 514
graduates across the three major groups was collected according to their
relative employment share. All the questionnaires have been conducted by
face-to-face interviews.
3.1. The Measurement of Three Types of Job Mismatch
Regarding qualification-job mismatch, the empirical work so far has
relied on the three methods to measure required qualification. First,
the Job Analysts (JA) Method (iObjective Approach), in which the
professional job analysts grade the jobs and recommend the minimum
educational requirements for a certain job [Battu, et al. (2000)].
Second method refers to Self Assessment (Subjective approach), where
workers are asked directly to give information on the minimum
educational requirements for their current job or whether they are
mismatched or not [Alba (1993)]. The third method 'Realised match
(RM)' measures the degree of qualification-job mismatch by two
variables; years of schooling and occupation. The distribution of
education is calculated for each occupation; employees who depart from
the mean by some ad-hoc value (generally one) standard deviation are
classified as mismatched workers [Verdugo and Verdugo (1989) and Ng
(2001)].
This study has measured qualification-job mismatch by all the three
methods, which are job analyst (JA), worker self assessment (WSA), and
realised match (RM) on the basis of SEG 2010 dataset. The attained
education (number of completed years) has been used as a measure of
qualification; while the required qualification (education) has also
been measured in years. For the JA method, the required level of
qualification in years has been measured by questioning the sampled
graduates "In your opinion, what level of formal education (years)
and experience (years) is demanded by your employer/organisation to get
the job like yours?" For the WSA approach, graduates were asked
"In your opinion, how much formal education (years) and experience
(years) is required to perform your current job well?" By comparing
the attained qualification and required qualification, the graduates
have been classified into three categories; overqualified,
under-qualified and matched graduates.
For the third RM measure, the required qualification has been
measured on the basis of two variables; completed years of schooling and
occupations. The mean years of schooling at two-digit occupational
classification has been used as a measure of required qualification by
assuming that the graduates working in similar occupation require the
same level of qualification. The qualification-job mismatch has been
estimated by comparing the attained and required qualification with
(+/-) one standard deviation of the mean. (6) Graduates with attained
qualification greater and less than one standard deviation were defined
as over-qualified and under-qualified graduates, respectively. The
middle range; within +/- of one standard deviation comprised of the
matched workers.
Following Chevalier (2003), a measure of qualification-job mismatch
and occupation-satisfaction has also been adopted to capture the
idiosyncratic characteristics by segregating the over-qualified
graduates into two categories; those over-qualified who are satisfied
over their mismatch are defined as apparently over-qualified, whereas
those who are dissatisfied are genuinely over-qualified. (7)
Skill is a broad signal of human capital because it assimilates the
other constituents of human capital (skills, experience) and also the
formal qualification/education. The attained skills possessed by the
workers, may be lower or higher than the required skills in their
prospective jobs, known as mismatch in skill. Majority of the studies
have used formal education as the proxy of skill; (8) however, the later
studies have criticised it as it is difficult to quantify the extent of
this skill [Jim and Egbert (2005); Lourdes, et al. (2005)]. The two
measurement approaches of skill mismatch have emerged from the
literature; majority of the studies have used the subjective approach,
based on worker's perception [Green and McIntosh (2002); Lourdes,
et al. (2005)], while some studies have used the specific approach by
measuring the various specific attained skills possessed by the workers
and the required skills in their current jobs [Jim and Egbert (2005);
and Chevalier and Lindley (2006)].
The ongoing study has followed the specific approach to measure
skill mismatch where initially, the level of nine specific attained and
required skills have been estimated in SEG survey on five-point scale,
ranging from 1 'not at all' to 5 'a lot'. These nine
skills are; supervisory skills, English writing skills, English speaking
skills, numeracy skills, teamwork skills, management skills, computer
skills, research skills and time management skills. Through Principal
Component Analysis (PCA) method, the weights has been estimated on
attained skills and required skills on the basis of mean required level
of nine skills by assuming that the workers in same occupations at
two-digit occupational coding require the similar types of skills in
their jobs. The skill mismatch has been estimated by comparing the
attained skill index and required skill index with (+/-) 0.08 standard
deviation (SD) of the mean (0.075 SD for SEG weights). (9) The graduates
with attained skills more or less than required skills by 0.08 standard
deviation were defined as over-skilled and under-skilled, respectively.
The middle range comprises the skill matched graduates. For detail
methodology along with questions on attained and required skills, see
Farooq (2011).
The field of study and job mismatch analyses the level of match
between the individual's field of study and his/her features of the
job. The existing three studies have adopted both subjective and
education-occupation combination to measure the field of study and job
mismatch [Jim and Robert (2004); Robst (2007) and Martin, et a/.(2008)].
The ongoing study has estimated the field of study and job mismatch by
subjective approach with the question: 'how much is your current
job relevant to your areas of education?' The four possible options
were; irrelevant field of study, slightly relevant, moderately relevant
and completely relevant field of study.
3.3. Impact of Job Mismatch on Earnings: Methodology
The specification to estimate the impact of job mismatch on
earnings revolves around the standard Mincer earning equation [Mincer
(1974)], which itself was originated to measure Becker's human
capital theory (1964). The standard Mincer earnings equation is
generally written as:
Ln[y.sub.i] = [[delta].sub.0] + [[delta].sub.1] [Year_School.sub.i]
+ [delta]'[X.sub.ki] + [mu] (1)
Where, Ln[y.sub.i] is natural log of monthly wages, year of
schooling measure the impact of attained qualification on earning while
X, represents the vector of all independent control variables related to
personal characteristics and human capital characteristics. In contrast
to the HCT, one can measure the Job Competition Theory [Thurow (1975)]
by replacing the required qualification with attained qualification in
Equation 1.The job assignment theory provides the framework to analyse
the impact of job mismatch on earning by adding over-qualification and
under-qualification. Two types of model specifications have been applied
so far in the literature to measure the impact of qualification-job
mismatch on earnings as given in the following two equations:
Ln [y.sub.i] = [[alpha].sub.0] + [[alpha].sub.1] [Q.sup.r.sub.1] +
[[alpha].sub.2] [Q.sup.o.sub.1] + [[alpha].sub.3] [Q.sup.u.sub.1] +
[alpha]'[X.sub.i] + [[epsilon].sub.i] (2)
Ln [y.sub.i] = [[beta].sub.0] + [[beta].sub.1] [Year_school.sub.1]
+ [[beta].sub.2][D.sup.uq.sub.i] + [beta]'[D.sup.uq.sub.i] +
[beta]'[X.sub.i] + [[epsilon].sub.i] (3)
In Equation 2, the years of required qualification ([Q.sup.r]),
years of over-qualification ([Q.sup.0]) and years of under-qualification
([Q.sup.u]) have been used as explanatory variables to analyse the
impact on earnings. In Equation 3, the former methodology has been
modified by taking dummy variables of over-qualification ([D.sup.oq])
and under-qualification ([D.sup.uq]). The core difference between the
two approaches is when one measures the qualification-job mismatch in
terms of years, then the coefficients of over-qualification and
under-qualification should be compared with those workers who are
matched but on the same jobs; whereas, in dummy specification, the
over-qualified and under-qualified graduates have been compared with
those who have same qualification but on matched jobs. As this study has
targeted the graduate employees, therefore, being limited variation in
years of over-qualification and years of under-qualification variables,
the second approach has been adopted. Another advantage of using the
second approach is that it has the capability to split
over-qualification ([D.sup.uq.sub.i]) variable into genuinely
over-qualified ([D.sup.ouq]) and apparently over-qualified ([D.sup.ouq])
category to capture the heterogeneity among the skills of graduates,
thus resulting in the following equation;
Ln [y.sub.i] = [[beta].sub.0] + [[beta].sub.1] [Year_school.sub.i]
+ [[beta].sub.2][D.sup.ogq.sub.i] + [[beta].sub.3][D.sup.oaq.sub.i] +
[[beta].sub.4][D.sup.uq.sub.i] + [beta]'[X.sub.ki] + [[mu].sub.i].
(4)
In the light of Mincerian earning equation, the following equation
has been used to measure the impact of skill mismatch on graduates'
earnings where [os.sub.i] and [us.sub.i] are dummy variables for
over-skill and under-skill for graduate /;
Lnyi = [[beta].sub.0] + [[beta].sub.1] [Year_school.sub.i] +
[[beta].sub.2][os.sub.i] + [[beta].sub.3][us.sub.i] +
[beta]'[X.sub.i] + [[epsilon].sub.i] (5)
The following equation has been used to measure the impact of field
of study and job mismatch on graduates' earnings where [sr.sub.i],
[mr.sub.i] and [cr.sub.i] represent the three dummies for weakly
relevant, moderately relevant and completely relevant field of study to
the current job:
Lnyi = [[beta].sub.0] + [[beta].sub.1] [Year_school.sub.i] +
[[beta].sub.2][sr.sub.i] + [[beta].sub.3][mr.sub.i] +
[[beta].sub.0][cr.sub.i] + [beta]'[X.sub.i] + [[epsilon].sub.i] (6)
4. RESULTS
4.1. Incidences of Job Mismatch
The estimates in Table 1 show that the incidence of
qualification-job mismatch varies by the three measures, which are
worker's self assessment (WSA), job analysis (JA) and realised
match (RM) method. Both the WSA and JA show that the level of
over-qualification and under-qualification are close to each other as
compared to the RM measure. The close estimates of over-qualification by
WSA and JA approach suggest that graduates have not overstated or
understated the qualification requirements. These estimates are
consistent with the earlier findings that RM method reports a lower
incidence of over-qualification as compared to the WSA and JA methods
[Meta-analysis of Groot and Maassen (2000) and McGuinnes (2006)].High
statistical relation was found between WSA and JA while poor
relationship was found with RM of both JA and WSA. (10)
To get a realistic picture, the assumption of'homogeneity in
skills of workers who hold the same qualification level', has been
relaxed by segregating the over-qualified workers into 'apparently
over-qualified' and 'genuinely over-qualified' on the
basis of occupation-satisfaction approach. Table 2 shows that under WSA
and JA approaches, about 57 to 63 percent of the over-qualified
respondents in non-graduate jobs are not too dissatisfied with their
mismatch, therefore, they are defined as apparently over-qualified
graduates and the rest (37 percent to 43 percent) who are dissatisfied,
are defined as genuinely over-qualified graduates. The issue of
heterogeneity of jobs is now clear as the genuinely and apparently
over-qualified graduates are not similar in skill possession. These
results are consistent with the earlier studies, which have captured the
issue of heterogeneity [Chevalier (2003); Chevalier and Lindley (2006)].
The results over skill mismatch have been reported in Table 3,
which shows that more than one-fourth of the graduates are mismatched in
skill either in terms of being over-skilled or in terms of being
under-skilled. The phenomenon of 'matched graduates' is
considerably higher among males (73 percent-74 percent) than among
females (67 percent). A lesser proportion of female graduates are
under-skilled, while, there are more over-skilled female graduates. It
reflects the scenario of relatively more under-utilisation of
females' skills in their jobs in Pakistan.
The results for the field of study and job mismatch have been
reported in Table 4, which shows that 11 percent of the graduates
consider that their current jobs are totally irrelevant to their studied
field of discipline, while another 14 percent reported their jobs are
slightly relevant, followed by the moderately relevant with 38 percent
and completely relevant with 37 percent. An important information is
that the female graduates are facing more field of study and job
mismatch than the male graduates as one-third of the female graduates
are mismatched falling in either irrelevant or weakly relevant category;
however, less than one-fourth of the male graduates are falling in these
first two categories (Table 3). See Farooq (2011) whether the formal
education is good proxy of skill or not?
4.2. Impact of Job Mismatch on Graduates' Earnings
In the light of Equations 3 and 4, Table 5 reports the estimated
results of qualification-job mismatch where model 1 and model 2 estimate
the impact of qualification-job mismatch on graduates' earning by
WSA and JA approach. In model 3 and model 4, the over-qualified
graduates have further been split into genuinely over-qualified and
apparently over-qualified. The exponential criteria has been adopted to
calculate the percentage impact of indicator variables. The residuals of
all the 4 models have been reported in Appendix Figures 1 to Figure 4,
which are normally distributed, sugesting that the t-stat values are
reliable. The coefficients of over-qualification in model 1 and model 2
show that over-qualified graduates face 30 percent to 37 percent of wage
penalty under different approaches (WSA and JA). The results are in line
with existing studies of qualification-job mismatch, which support the
job assignment model [Sattinger (1993)] that both individual and job
characteristics determine the level of earnings. These results are also
in the line with previous studies that both WSA and JA yield consistent
results, with the overestimation by WSA approach [McGoldrick and Robst
(1996); Battu, et al. (2000); Groot and Maasen (2000)]. After
controlling the heterogeneity in model 3 and model 4 by splitting the
over-qualified graduates into 'genuine' and
'apparent' category, the penalty for over-qualification is
still statistically significant with less penalty to apparently
over-qualified (20 percent to 26 percent) and more to the genuinely
over-qualified graduates (49 percent to 53 percent) under WSA and JA
approaches. The coefficient of under-qualification is not significant in
all the models. These results are consistent with the earlier studies
that the genuinely overqualified face more wage penalties as compared to
apparently over-qualified [Chevalier (2003); Chevalier and Lindley
(2006)].
Regarding the other control variables, all the models show that the
male graduates are likely to earn 10 percent to 12 percent more than the
female graduates, consistent with earlier studies conducted in Pakistan
[Sabot (1992); Nazli (2004); Nasir (2002, 2005) and many others)]. The
significant coefficients for education and experience show the
importance of human capital accumulation as the graduates with more
education and experience have a positive rate of return on it. Regarding
the quality of institution from where the graduates have obtained their
highest degree, the graduates who got their education from distance
learning institutes earn about 32 percent less than those who got their
education from the university. The foreign degree/diploma holders
graduates earn about 20 to 23 percent more than the locally educated.
These differences reflect the heterogeneity of education, which in turn
is generating the wage differences among the graduates.
Regarding the labour market characteristics, a wage differential
exists between government and private organisations where graduates in
the government sector earn less than the private sector. Tenure with the
current job also has a strong influence on graduates' earnings, as
the graduates who have been in the current job between two to four years
earn about 20 percent to 22 percent more and the graduates with more
than four years in the current job earn 30 percent to 32 percent more
than those who have tenure up to one year (Table 5).
Following Equations 5 and 6, the results are given in Table 6 where
model 5 measures the impact of skill mismatch on earnings, while model
6measures the impact of field of study and job mismatch. The residuals
of both models have been reported in Appendix Figure 5 to Figure 6. The
results about the impact of skill mismatch on graduates' earnings
in model 5 show that over-skilled graduates face 20 percent wage
penalties and under-skilled get 16 percent wage premium as compared to
those who have the same level of education and on matched jobs.
Regarding the under-skilled, the findings of this study are different
from the studies of Lourdes, et al. (2005) in which the under-skilled
workers face wage penalties; however, the estimates of this study are in
the right direction that under-skilled graduates get wage premium when
compared with the matched workers. These results are consistent with the
earlier studies, which indicate that skill mismatch leads to wage
differential among the workers [Green and McIntosh (2002); Lourdes, et
al. (2005); Di-Pietro and Urwin (2006)].
In the last model, the estimates show that the moderate field of
study and job matched and complete field of study and job matched
graduates earn significantly more by 23 percent and 20 percent
respectively compared to those who have irrelevant field of study in
their current jobs. These results are in line with existing studies
showing that a good match between the field of study and the current job
improves the level of earnings [Robst (2007); Martin, et al. (2008);
Domadenik, et al. (2013)].
Regarding gender, the estimates support the initial results as
mentioned in Table 5 that male graduates, on average, earn 11 percent
more than the female graduates. Similarly, education and experience have
a significant impact on graduates' earnings with 10 percent and 3
percent per year, respectively. The graduates with foreign diploma earn
more than the locally educated graduates (Table 6).
5. CONCLUSIONS AND POLICY IMPLICATIONS
The main focus of this study is to estimate the three types of job
mismatches and analysing the pecuniary consequences of job mismatch. The
present study has found that the choice of measurement method has a
significant effect on the incidences of qualification-job mismatch.
Overall 31-37 percent of the graduates are facing the qualification-job
mismatch either falling in over-qualification or under-qualification
category. Similarly, more than one-fourth of the graduates are
mismatched in skill either in terms of being over-skilled or in terms of
being under-skilled. The phenomenon of 'matched graduates' is
considerably higher among males than among females. An important
information is that the female graduates are facing more field of study
and job mismatch than the male graduates as one-third of the female
graduates are mismatched falling in either irrelevant or weakly relevant
category; however, less than one-fourth of the male graduates are
falling in these two categories.
This study has examined the impact of all the three types of job
mismatches on graduates' earnings and found that the over-qualified
graduates face 30 to 37 percent wage penalty under different approaches.
After controlling skill heterogeneity, the penalty for over-education is
still significant with fewer penalties to apparently over-qualified and
more penalties to genuinely over-qualified. The over-skilled graduates
face wage penalties and the under-skilled get wage premium as compared
to the matched workers. A good field of study and job match also improve
the wages of graduates. Overall these results do not support the Human
Capital Theory. However, this study cannot necessarily reject the Human
Capital Theory on the basis of cross-sectional dataset as the mismatch
phenomenon might be temporary. The results of this study support the Job
Assignment Theory [Sattinger (1993)] as both the individual and job
characteristics are determining the levels of job mismatch and wages.
Our findings lead to the following policy implications and
recommendations primarily in two areas; reforms in human resource
development and labour market institutions:
* The incidences of various types of job mismatches especially the
skill mismatch suggest the need for better quality of education and
skills by ensuring the equality of skills and rightly demanded skills
across the institutes and regions. The phenomenon of field of study and
job mismatch suggests the close coordination among the various demand
and supply side stakeholders of the labour market for better
understanding of issues in order to formulate the right policies.
* The rapid enrolment at higher education level with limited labour
demand suggests to implement entrepreneurial reforms both in educational
institutes and in the labour market to absorb this educated influx.
Females should receive a special focus in such policies, which would not
only raise their participation but also provide them the entrepreneurial
opportunities.
* Some tracer type studies or panel studies are required for a
better understanding of employment patterns and skills demanded by the
various sectors and occupations. It would not only guide the planners
and enrolled youths about the labour market opportunities and type of
skills needed, but also would help to project future educational needs.
* There is a need to improve the Labour Force Survey (LFS)
questionnaire for skill assessment and job mismatches. A module about
the history of employment may also be made part of the LFS. Additional
research is of course needed to estimate the timing and depth of job
mismatch, productivity losses and direct and indirect hiring and firing
costs to firms due to job mismatch.
[APPENDIX FIGURE 1 OMITTED]
[APPENDIX FIGURE 2 OMITTED]
[APPENDIX FIGURE 3 OMITTED]
[APPENDIX FIGURE 4 OMITTED]
[APPENDIX FIGURE 5 OMITTED]
[APPENDIX FIGURE 6 OMITTED]
REFERENCES
Afzal, Mohammad (2011) Microeconometric Analysis of Private Returns
to Education. Pakistan Economic and Social Review 49:1, 39-68.
Akbari, Ather H. and N. Muhammad (2000) Educational Quality and
Labour Market Performance in Developing Countries: Some Evidence from
Pakistan. The Pakistan Development Review 39:4, 417-439.
Alba-Ramirez, A. (1993) Mismatches in Spanish Labour Market:
Overeducation? The Journal of Human Resources 27;2, 259-278.
Aslam, Monazza (2005) Rates of Return to Education by Gender in
Pakistan. Global Poverty Research Group.
Battu, H., C. Belfield, and P. J. Sloane (1999) Overeducation among
Graduates: A Cohort View. Education Economics 7, 21-38.
Battu, H., C. Belfield, and P. J. Sloane (2000) How Well Can We
Measure Graduate Overeducation and Its Effects? National Institute
Economic ReviewM 1, 82-93.
Bauer, T. K. (2002) Educational Mismatch and Wages: A Panel
Analysis. Economics of Education Review 21,221-229.
Becker, Gary S. (1964) Human Capital: A Theoretical and Empirical
Analysis with Special Reference to Education. New York, Columbia
University Press.
Budria, S. and A. I. Moro-Egido (2008) Education, Educational
Mismatch, and Wage Inequality: Evidence for Spain. Economics of
Education Review 27:3, 332-341.
Chaudhry, Imran Sharif, Muhammad Zahir Faridi, and Sabiha Anjum
(2010) The Effects of Health and Education on Female Earnings: Empirical
Evidence from District Vehari. Pakistan Journal of Social Science 30:1,
109-124.
Chevalier, A. (2003) Measuring Over-education. Economics 70.
509-531.
Chevalier, A. and J. Lindley (2006) Over-education and the Skills
of U.K Graduates. IZA (Discussion Paper No. 2447).
Cohn, E. and S. P. Khan (1995) The Wage Effects of Overschooling
Revisited. Labour Economics 2, 67-76.
Di-Pietro, G. and P. Urwin (2006) Education and Skills Mismatch in
the Italian Graduate Labour Market. Applied Economics 38:1, 79-93.
Dolton, P. and M. Siles (2003) The Determinants and Consequences of
Overeducation. In Buchel, de Grip and Mertens (eds.) Overeducation in
Europe, (pp. 189-217). Cheltenham: Edward Elgar.
Dolton, P. and A. Vignoles (2000) The Incidence and Effects of
Overeducation in the UK Graduate Labour Market. Economics of Education
Review\9\2, 179-198.
Domadenik, P. Dasa Farcnik and Francesco Pastore (2013) Horizontal
Mismatch in the Labour Market of Graduates: The Role of Signalling. (IZA
DP No. 7527).
Duncan, G. J. and S. D. Hoffmann (1981) The Incidence and Wage
Effects of Overeducation. Economics of Education Review 1:1, 75-86.
Farooq, Shujaat (2011) The Utilisation of Education and Skills:
Incidence and Determinants among Pakistani Graduates. The Pakistan
Development Review 50:3, 219-244.
Freeman, R. (1976) The Overeducated American. New York: Academic
Press.
Frenette, M. (2004) The Overqualified Canadian Graduate: The Role
of the Academic Program in the Incidence, Persistence, and Economic
Return to Overqualification. Economics of Education Review 23, 29-45.
Ghayur, S. (1989) Educated Unemployed in Pakistan: Estimates of
Imbalances in the Current Flows. The Pakistan Development Review 28, 4:
603-613.
Gill, A. M. and E. J. Solberg (1992) Surplus Schooling and
Earnings: A Critique. The Journal of Human Resources 27:4, 683-689.
Green, F. and S. McIntosh (2002) Is there a Genuine
Underutilisation of Skills Amongst the Overqualified? London School of
Economics.
Groeneveld, S. and J. Hartog (2004) Overeducation, Wages and
Promotions within the Firm. Labour Economics 11, 701-714.
Groot, W. (1996) The Incidence of and Returns to Overeducation in
the UK. Applied Economics 28, 1345-1350.
Groot, W. and van den B. Maasen (1997) Allocation and the Returns
to Overeducation in the UK. Education Economics 50:2, 169-183.
Groot, W. and van den B. Maasen (2000) Overeducation in the Labor
Market: A Meta-Analysis. Economic of Education Review 19, 149-158.
Haque, Nadeem Ul, G. M. Arif, and N. Iqbal (2007) Growth, Poverty
and Social Outcomes in Pakistan. A Study Prepared for DFID, Islamabad.
Hartog, J. (2000) Overeducation and Earnings: Where are we, Where
Should we go? Economics of Education Review 19, 131-147.
Hausman, R. D. Rodrik, and A. Velasco (2005) Growth Diagnostics.
John F. Kennedy School of Government, Harvard University.
Jim, A., and de V. Robert (2004). Determinants of Skill Mismatches:
The Role of Learning Environment, the Match between Education and Job
Working Experience. Paper presented at TLM.NET Conference: Quality in
Labour Market Transitions; a European Challenge. November 2004 Royal
Academy of Sciences Amsterdam.
Jim, A., ROA and de W. Egbert (2005) What do Educational Mismatches
Tell us about Skill Mismatches? A Cross Country Analysis. Paper for the
Seminar: European Labour Market of Higher Education Graduates: Analysis
of the CHEERS Project Segovia.
Kiker, B., M. Santos, and M. de Oliveiria (1997) Overeducation and
Undereducation: Evidence for Portugal. Economics of Education Review
16:2, 111-125.
Leuven, Edwin and Hessel Oosterbeek (2011) Overeducation and
Mismatch in the Labour Market. (1ZA DP No. 5523).
Lourdes, Badillo-Amador, A. Garcia-Sanchez, and L. E. Villa (2005)
Mismatches in the Spanish Labor Market: Education vs. Competences Match.
International Advances in Economic Research 11, 93-109.
Lourdes, Badillo-Amador, and L. E. Vila-Luis (2013) Education and
Skill Mismatches: Wage and Job Satisfaction Consequences. International
Journal of Manpower 34:5, 416-428.
Martin, N., I. Persson, and Dan-Olof Rooth (2008)
Education-Occupation Mismatch: Is there an Income Penalty? (IZA
Discussion Paper No. 3806).
McGoldrick, K. and J. Robst (1996) Gender Differences in
Overeducation: A Test of the Theory of Differential Overqualification.
American Economic Review 86, 280-304.
McGuinness, S. (2003) Private Sector Post Graduate Training and
Graduate Under-Employment. Evidence from Northern Ireland. International
Journal of Manpower 23:6, 527-541.
McGuinness, S. (2006) Overeducation in the Labour Market. Journal
of Economic Surveys 20:3, 387-418.
Mincer, J. (1974) Schooling, Experience and Earnings. New York:
Columbia University Press.
Nasir, Mueen Zafar (2002) Returns to Human Capital in Pakistan: A
Gender Disaggregated Analysis. The Pakistan Development Review 41:1,
1-28.
Nasir, Mueen Zafar (2005) An Analysis of Occupational Choice in
Pakistan: A Multinomial Approach. The Pakistan Development Review 44: 1
(Part 11), 57-79.
Nazli, Hina (2004) The Effects of Education, Experience and
Occupation on Earnings: Evidence from Pakistan. The Lahore Journal of
Economics 9,1-30.
Ng, Y. C. (2001) Overeducation and Undereducation and Their Effect
on Earnings: Evidence from Hong Kong, 1986-1996. Pacific Economic Review
6:3, 401-418.
Pakistan, Government of (2011) Pakistan Employment Trends, Ministry
of Labour, Manpower and Overseas Pakistanis, Labour Market Information
and Analysis Unit, Islamabad.
Pakistan, Government of (2013) Labour Force Survey, 2010. Federal
Bureau of Statistics, Islamabad.
Pakistan, Government of (2008) Pakistan Employment Trends Youth,
Ministry of Labour, Manpower and Overseas Pakistanis, Labour Market
Information and Analysis Unit, Islamabad.
Psacharopoulos, G. and Patrinos (2002) Returns to Education: A
Further International Update and Implications. The Journal of Human
Resources 20:4.
Qayyum, Abdul, et al. (2007) Growth Diagnostics in Pakistan. (PIDE
Working Paper 2006-07).
Robst, J. (2007) Education and Job Match: The Relatedness of
College Major and Work. Economics of Education Review 26:4, 397-407.
Rumberger, R. W. (1987) The Impact of Surplus Schooling on
Productivity and Earnings. Journal of Human Resources 22:1, 24-50.
Sabot, R. (1992) Human Capital Accumulation in Post-green
Revolution Rural Pakistan: A Progress Report. The Pakistan Development
Review 31:4, Part 1.
Sattinger, M. (1993) Assignment Models of the Distribution of
Earnings. Journal of Economic Literature 31:2, 831-880.
Schultz, T. W. (1962) Investment in Human Capital. American
Economic Review 51:1, 1-17.
Shabbir, T. (1993) Productivity-Enhancing vs. Credentialist Effects
of Schooling in Rural Pakistan. International Food Policy Research
Institute, Islamabad.
Sicherman, N. (1991) Overeducation in the Labour Market. Journal of
Labour Economics 9:2.
Sloane, P. J., H. Battu, and P. T. Seaman (1999) Overeducation,
Undereducation and the British Labour Force. Applied Economics 31:11,
1437-1453.
Thurow, L. C. (1975). Generating Inequality. New York: Basic Books.
Tsang, M. C. (1987) The Impact of Underutilisation of Education and
Productivity: A Case Study of the U.S. Bell Companies. Economics of
Education Review 6:2 239-254.
Verdugo, R. R. and N. T. Verdugo (1989) The Impact of Surplus
Schooling on Earnings: Some Additional Findings. Journal of Human
Resources 24:4, 629-643.
Zahid, Gulnaz (2014) Role of Career Education Advisor/Expert and
Teaching Quality in Student Employability Skills as the Outcome of
Higher Education. Mediterranean Journal of Social Sciences 5:27.
(1) For U.K, 12 percent by Dolton and Vignoles (2000), 18 percent
by Dolton and Silles (2003), 23.2 percent by Chevalier and Lindley
(2006). For U.S, 13 percent by Verdugo and Verdugo (1989), 11 percent by
Cohn and Khan (1995). For Holland, 26 percent by Groot (1996), 8 percent
in Kiker, et al. (1997) for Portugal and 27 percent in Budria and Edigo
(2007) for Spain.
(2) 61.2 percent were considered vulnerable, meaning "at risk
of lacking decent work" in 2012-13 [Pakistan (2013)].
(3) Shabbir (1993), Nasir (2002, 2005), Akbari, et al. (2000),
Nazli (2004), Aslam (2005), Chaudhary, et al. (2010), Afzal (2011) and
many others.
(4) Govemment of Pakistan (2003-04) "Thirteenth Census of
Federal Government Civil Servants". Pakistan Public Administration
Research Centre, Management Services Wing, Establishment Division,
Islamabad.
(5) Government of Pakistan (2007-08) "Annual Statistical
Bulletin of Federal Government". Pakistan Public Administration
Research Centre, Management Services Wing, Establishment Division,
Islamabad.
(6) +/- One standard deviation was used as the actual mean
deviation of the difference of the attained education and the required
education was 0.989, close to one.
(7) Job satisfaction has been measured at five point Likert scale
range from very dissatisfied to very satisfied. For apparently
over-qualified workers, range 1 (very dissatisfied) and range 2
(dissatisfied) were used while for genuinely over-qualified workers
range 3 to 5 have been used.
(8) As Battu, el al. (1999), Frenette (2004), Groot (1996) and Ng
(2001) did.
(9) Standard deviation has been calculated after comparing the both
attained and required skill index.
(10) Parametric t-test and spearman rank correlation tests were
applied.
Shujaat Farooq <
[email protected]> is Research Economist,
Pakistan Institute of Development Economics (PIDE), Islamabad.
Author's Note: The author completed a PhD in Economics at PIDE
in 2011. This paper is part of his doctbral dissertation. He is grateful
to his supervisors, Dr G. M. Arif, Joint Director of PIDE, and Dr Abdul
Qayyum, Joint Director of PIDE, for their valuable suggestions and
guidance.
Table 1
The Level of Qualification-Job Mismatch by Various Measures (%)
Measures Matched Under-qualified Over-qualified
WSA Method 65.4 9.9 24.7
JA Method 69.5 4.5 26.1
RM Method 63.4 21.6 15.0
Table 2
The Level of Genuine and Apparent Over-qualification (%)
Education-Job Mismatch WSA Approach JA Approach RM Approach
Matched 65.4 69.5 63.4
Under-qualified 9.9 4.5 21.6
Genuinely Over-qualified 10.7 9.7 4.7
Apparently Over-qualified 14.0 16.3 10.3
Table 3
The Distribution of Respondents by the Level of Skill Mismatch (%)
Matched Graduates Under-skilled Over-skilled
Female 66.7 11.1 22.2
Male 72.8 13.9 13.4
Both Sexes 71.8 13.4 14.8
Table 4
% Distribution of the Respondents by Field of Study and
Job Mismatch
Level of Mismatch Female Male Total
Irrelevant 14.8 10.6 11.3
Slightly Relevant 18.5 12.9 13.8
Moderately Relevant 33.3 39.3 38.3
Completely Relevant 33.3 37.2 36.6
Table 5
The Impact of three Types of Job Mismatch on Graduates'
Earnings--SEG, 2010
Model I
WSA-I
Regressor Coeff. St. Err.
Over-qualification -0.367 * 0.060
Under-qualification -0.051 0.079
Over-qualification genuine -- --
Over-qualification apparent -- --
Education 0.136 * 0.024
Experience 0.025 * 0.009
Experience square -0.017 * 0.008
Sex (male=l) 0.113 ** 0.063
Marital status (married=l) 0.118 * 0.06
Foreign diploma (yes=l) 0.226 * 0.087
Type of institution
(university as ref.)
College -0.050 0.068
Distance learning -0.282 * 0.084
Organisation of job (govt =1) -0 049 ** 0.03
Tenure (up to 1 year as ref.)
1 to 2 year 0.019 0.082
2 to 4 year 0.212 * 0.077
More than 4 year 0.322 * 0.090
Constant 7.430 * 0.408
F-Stat 17.99
R-square 0.5759
N
Model 2
JA-1
Regressor Coeff. St. Err.
Over-qualification -0.295 * 0.061
Under-qualification -0.051 0.111
Over-qualification genuine -- --
Over-qualification apparent -- --
Education 0.138 * 0.025
Experience 0.027 * 0.01
Experience square -0.016 * 0.009
Sex (male=l) 0.118 ** 0.063
Marital status (married=l) 0.117 ** 0.061
Foreign diploma (yes=l) 0.209 * 0.088
Type of institution
(university as ref.)
College -0.07 0.069
Distance learning -0.279 * 0.086
Organisation of job (govt =1) -0.050 ** 0.03
Tenure (up to 1 year as ref.)
1 to 2 year -0.01 0.083
2 to 4 year 0.195 * 0.078
More than 4 year 0.305 * 0.091
Constant 7.395 * 0.415
F-Stat 17.17
R-square 0.5644
N 514
Model 3
WSA-I I
Regressor Coeff. St. Err.
Over-qualification -- --
Under-qualification -0.044 0.078
Over-qualification genuine -0.532 * 0.081
Over-qualification apparent -0.265 * 0.068
Education 0.139 * 0.024
Experience 0.024 * 0.009
Experience square -0.017 * 0.008
Sex (male=l) 0.114 ** 0.062
Marital status (married=l) 0.118 * 0.06
Foreign diploma (yes=l) 0.207 * 0.086
Type of institution
(university as ref.)
College -0.055 0.067
Distance learning -0.292 * 0.084
Organisation of job (govt =1) -0.045 ** 0.027
Tenure (up to 1 year as ref.)
1 to 2 year 0.007 0.081
2 to 4 year 0.205 * 0.076
More than 4 year 0.306 * 0.089
Constant 7.409 * 0.404
F-Stat 18.06
R-square 0.5840
N
Model 4
JA-II
Regressor Coeff St. Err.
Over-qualification -- --
Under-qualification -0.044 0.110
Over-qualification genuine -0.487 * 0.085
Over-qualification apparent -0.203 * 0.067
Education 0.142 * 0.025
Experience 0.025 * 0.009
Experience square -0.016 * 0,009
Sex (male=l) 0.121 ** 0.063
Marital status (married=l) 0.120 * 0.061
Foreign diploma (yes=l) 0.203 * 0.087
Type of institution
(university as ref.)
College -0.067 0.068
Distance learning -0.287 * 0.085
Organisation of job (govt =1) -0.048 ** 0.030
Tenure (up to 1 year as ref.)
1 to 2 year -0.017 0.082
2 to 4 year 0.181 * 0.078
More than 4 year 0.291 * 0 091
Constant 7.366 * 0.411
F-Stat 17.30
R-square 0.5735
N
* Denotes significant at 5 percent, ** denotes
significant at 10 percent.
Table 6
The Impact of three Types of Job Mismatch on Graduates'
Earnings--SEG, 2010
Model 5
Regressor Skill Mismatch
Coeff St. Err.
Over-skill -0.195 * 0.066
Under-skill 0.155 * 0.069
Weak relevance/irrelevant -- --
Moderate relevance/irrelevant -- --
Complete relevance/irrelevant -- --
Education 0.102 * 0.023
Experience 0.026 * 0,01
Experience square -0.017 * 0.008
Sex (male=l) 0.102 ** 0.063
Marital status (married=l) 0.103 ** 0.062
Foreign diploma (yes=l) 0.194 * 0.089
Type of institution (university as ref.)
College -0.073 0.069
Distance learning -0.276 * 0.086
Organisation of job (govt.=l) -0.056 ** 0.03
Tenure (up to 1 year as ref.)
1 to 2 year -0.018 0.084
2 to 4 year 0.197 * 0.079
More than 4 year 0.292 * 0.092
Constant 7.866 * 0.393
F-Stat 16.67
R-square 0.5572
N 514
Model 6
Regressor Field of study
Mismatch
Coeff. St. Err.
Over-skill -- --
Under-skill -- --
Weak relevance/irrelevant 0.115 0.09
Moderate relevance/irrelevant 0.228 * 0.083
Complete relevance/irrelevant 0.203 * 0.09
Education 0.102 * 0.024
Experience 0.029 * 0.01
Experience square -0,016 * 0.009
Sex (male=l) 0.099 ** 0.062
Marital status (married=l) 0.118 ** 0.062
Foreign diploma (yes=l) 0.218 * 0.09
Type of institution (university as ref.)
College -0.043 0.07
Distance learning -0.260 * 0.088
Organisation of job (govt.=l) -0.053 ** 0.031
Tenure (up to 1 year as ref.)
1 to 2 year 0.000 0.084
2 to 4 year 0.216 * 0.079
More than 4 year 0.298 * 0.093
Constant 7.735 * 0.409
F-Stat 15.75
R-square 0.55
N
* Denotes significant at 5 percent, ** Denotes significant
at 10 percent.
